To investigate the secondary organic aerosol (SOA) formation in Beijing, fine particle (PM2.5) samples were collected at an urban site (Peking University, PKU) and a rural site (Yufa) during CAREBEITING 2008 summer intensive field campaign. Several approaches were used to estimate SOA concentrations, including tracer-yield method, non-primary organic carbon method (receptor model, Chemical Mass Balance model), non-biomass burning water soluble organic carbon (WSOC) method and EC tracer OC/EC ratio method. To develop non-biomass burning WSOC method, simulation of typical Chinese biomass burning was conducted to obtain the important parameters. The average WSOC/OC ratio in Chinese biomass burning is 0.48 +/- 0.04. This ratio can be used in non-biomass burning WSOC method to estimate SOC of China. The results from different methods all indicated secondary formation has become major contributor to organic aerosols in Beijing, accounting for 50% or more of the total OC. A closure study was made by combination of CMB model and tracer-yield method. Five primary sources, including vegetative detritus, biomass burning, coal burning, gasoline engines and diesel engines, and four secondary organic aerosols derived from isoprene, alpha-pinene, beta-caryophyllene and toluene were apportioned. To the current knowledge, about 20%similar to 27% of the OC sources still remain unknown. Applicability of these SOC methods in China was tested by comparing the different methods. Tracer-yield method, CMB model and EC tracer OC/EC ratio method can be used to estimate SOC in China. However, non-biomass burning WSOC method can only be used to estimate water-soluble SOC. Uncertainty analysis was conducted to help researchers to determine the proper method to estimate SOC in China. Tracer-yield method underestimates total SOC, because it can only estimate SOCs from several precursors. Similarly, non-biomass burning WSOC method also underestimate SOC. Non-primary OC method overestimates SOC due to unapportioned primary OC, especially in urban area where the particle sources are complicated. The uncertainty of EC-tracer OC/EC ratio method is mainly from the primary OC/EC ratio. The largest overestimation and underestimation of single point value can be 54% and 64%.